Design of a Disease-Specific Master Protocol

Design of a Disease-Specific Master Protocol

Design of a Disease-Specific Master Protocol Roy Herbst, MD/PhD Yale Cancer Center

Modernizing Clinical Trial Process Some of the current challenges of drug development Difficulties in recruiting cancer patients to clinical trials Extensive bureaucratic processes required to initiate any clinical trial Lengthy regulatory review Modernizing trial process with innovative approaches and new clinical trial designs is of high importance. Use novel design strategies combined with biomarker testing to increase trial efficiency improve future phase III clinical trial designs

Phase II Adaptive Screening Trials BATTLE (Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination) Heavily pretreated refractory non-small cell lung cancer (NSCLC) Determined marker status of 11 biomarkers Randomized patients to four different agents Results were used to design two new BATTLE trials I-SPY 2 TRIAL (Investigation of Serial Studies to Predict Your Therapeutic Response with Imaging And molecular Analysis 2) Investigates neoadjuvant treatment of new drugs added to traditional chemotherapy in women with locally advanced breast cancer Designed to test multiple novel drugs and biomarkers over five-year timeframe May test up to five drugs simultaneously Add new drugs as existing drugs complete testing

Moving to a Multi-arm Registration Trial We propose another alternative to traditional trial design Multi-arm, multi-marker/drug master protocol Phase III trial Randomized, Controlled No adaptive randomization Multiple new therapies are tested simultaneously in a specific disease setting Designed to allow FDA approval of new therapeutics Assigns patients to experimental treatment vs standard-of-care control arm on the basis of specific biomarkers

Advantages of Master Protocol Multi-drug Registration Trial Design Grouping multiple studies reduces the overall screen failure rate Single master protocol will result in process and operational efficiency gains Provides consistency Trial infrastructure will be in place Bring safe and effective drugs to patients faster

Master Protocol Multi-Drug Trial Design Advanced NSCLC patients, multiple experimental drugs with matching predictive biomarker Interim analysis, PFS (and OS) endpoints Meets pre-specified endpoint Does not meet pre-specified endpoint Continue experimental arm Withdrawn from trial OS endpoint

Lung Cancer Example Squamous Cell Carcinoma Mutation Incidence Gene Event Type Frequency FGFR1 Amplification 20-25% FGFR2 Mutation 5% PIK3CA Mutation 9% PTEN Mutation-Deletion 18% CCND1 Amplification 8% CDKN2A Deletion/Mutation 45% PDGFRA Amplification-Mutation 9% EGFR Amplification 10% MCL1 Amplification 10% BRAF Mutation 3% DDR2 Mutation 4% cmet High copy-amplification 11% ERBB2 Amplification 2%

Speakers Roy Herbst, MD/PhD, Yale Cancer Center David Wholley, M.Phil., FNIH Eric Rubin, MD, Merck Lisa LaVange, PhD, FDA Jeff Abrams, MD, NCI Karen Arscott, DO, Lung Cancer Alliance Shakuntala Malik, MD, OHOP, FDA Acknowledgements Fred Hirsch, Vali Papadimitrakopoulou, David Gandara, Jim Doroshow, Gary Kelloff, Sonia Pearson-White, Keavan Anderson, Sharon Murphy, Lisa McShane

Design of a Disease-Specific Master Protocol Organizational infrastructure for multi-drug trials David Wholley FNIH

The I-SPY 2 trial tests multiple new breast cancer agents using biomarkers and an adaptive trial design ADAPT Paclitaxel + Trastuzumab ± New Drug A, B, or C AC (4 Cycles at 3 weeks each) HER2 (+) Randomize Surgery On Study Paclitaxel ± New Drug C, D or E (12 Weekly Cycles) AC (4 Cycles at 3 weeks each) HER2 ( ) Randomize Surgery Biopsy Blood Biopsy used for Biomarkers MRI MRI Biopsy MRI Biopsy Stratifying Biomarkers: Class I/II devices: HER2 (IHC, FISH) MammaPrint ER, PR IDE: MammaPrint44K Her2 (RPMA, 44K-microarray) MRI Blood Tissue

I-SPY 2 is being conducted as a large-scale public-private partnership managed through the FNIH Biomarkers Consortium Principal Investigators: Laura Esserman (UCSF) and Don Berry (MD Anderson) NCI FDA ~20 academic cancer centers Multiple pharmaceutical companies Contributing agents Funding Platform companies Laboratories Non-profit organizations Advocates Managed by FNIH and Quantum Leap Healthcare Collaborative FDA PUBLIC Enhance regulatory decision making Improve patient care and health Expedite drug development process INDUSTRY PARTNERING FOR THE PUBLIC HEALTH NIH Opportunities presented by science Clinical Research, science HEALTH PROFESSIONALS, SCIENTISTS

An Independent Agent Selection Committee chooses novel agents based on stringent criteria The IASC consists of 5-6 cancer drug development experts without current industry affiliations Phase I testing on candidate agents must be completed Agents must be compatible with standard paclitaxel therapy (no unacceptable additive toxicity) For HER2/neu-directed agents, compatible with paclitaxel plus trastuzumab therapy Known efficacy or rationale for efficacy in breast cancer Targets key pathways/molecules in breast cancer, but only one novel agent per target pathway will be accepted in the trial Fits strategic model for optimizing combinations of single/multiple molecular targeting drugs with or without standard chemotherapy Willingness of company to contribute agent and sufficient availability

I-SPY 2 Organizational Principles, Efficiencies FNIH holds a Master IND (developed with FDA and NCI) that incorporates testing of multiple agents FNIH negotiates and holds all contracts with sites, pharma companies, biomarker companies, and other entities FNIH and QuantumLeap provide centralized co-administration and project management for the trial A centralized IT infrastructure (based on cabig) ensures broad, timely dissemination of data and results Formal Data Access and Publication Guidelines ensure transparency and balance company and public health benefit Data and samples are made broadly available to the research community for follow-on research FNIH also serves as a trusted 3 rd party to manage data and intellectual property coming out of the trial, to maximize the public health benefit

FNIH acts as a trusted third party to ensure fair and appropriate licensing of new inventions arising from I SPY 2 1 Inventing Organizations grant exclusive licenses to new IP to FNIH Medical Center A Medical Center B 2 FNIH prosecutes and manages resulting patents Drug Co. A Laboratory C Drug Co. B Dx Co. C $$ 3 FNIH markets and licenses IP to interested parties 4 FNIH returns a fair share of royalties (less expenses) to Inventing Organizations

Design of a Disease-Specific Master Protocol Eric H. Rubin, MD Merck

Proposed Master Protocol Multi-drug Registration Trial Design Disease Model: Relapsed, refractory non-small cell lung cancer (NSCLC) Sponsor: A neutral third party e.g., CRO, academic coordinating center Able to establish appropriate firewall procedures Objective: To compare overall survival (OS) of biomarkerselected patients treated with standard of care (SoC) vs. experimental targeted therapy Standard of Care: Will be determined prior to trial initiation by the steering committee

Trial Design Drugs and Biomarkers The steering committee will evaluate each application to determine whether a drug/biomarker pair can enter the trial Drugs Ready to enter a phase III confirmatory trial Each drug must have clinical data demonstrating activity in a responsive patient group Patient group can be identified by assessment of biomarker in patient tumor biopsies

Trial Design Biomarkers and Screening Each compound s biomarker is based on analytically validated test/platform suitable for a pivotal trial This trial could use common screening platform that assays multiple biomarkers If predictive biomarker is in a CLIA-approved platform, it could be considered adequate for patient selection and randomization Would require Investigational Device Exemption (IDE) prior to trial start If new drug shows clinical benefit in selected patient population the biomarker could be analyzed and given FDA clearance

Advantages Use of a Multi-marker Platform Conserves tumor samples Testing protocols easier to standardize Sponsors would not be responsible for designing their own diagnostic Considerations Have not yet been used in registration trial The process would require close communication with the FDA to determine its applicability

Potential Study Design At entry, patients will receive a fresh core needle biopsy, with the tissue analyzed with appropriate assay (s) Experimental treatment A targets Marker A-positive tumors; Drug B targets Marker B-positive tumors Patients whose tumors are positive for marker A will be randomized to SoC vs drug A Patients whose tumors are positive for marker B will be randomized to SoC vs drug B Primary endpoint: Overall Survival (OS) Possible Interim Analysis when 30% of the OS events have occurred

Scenario: Two markers with no marker overlap; one drug per marker. Advanced NSCLC patients No marker overlap Core needle biopsy Stratify by biomarker expression Divide in 2 portions: 1) Clinical trial biomarker analysis 2) Save for future analysis Drugs A1 B1 A B A+B A B+ Randomize 1:1 Randomize 1:1 Drug A1 SoC Drug B1 SoC SoC PFS; OS PFS; OS

Master Protocol over Time Additional drug/biomarker combinations dropped and added to study Trial Start Patient Accrual/ Patient Screening for initial biomarkers Marker/Drug A treatment arm starts A HR=.84, Trial continues A 30% events Interim analysis for OS A 100% events OS analysis Master Protocol Trial Timeline Marker/Drug B treatment arm starts Marker/Drug C apply to trial Marker/Drug C reviewed by steering committee Marker/Drug C added to trial B 30% events Interim analysis for OS Marker/Drug C treatment arm starts B HR=0.92, Drug B dropped from trial C 30% events Interim analysis for OS

Case Study Example PIK3CA copy number increases or mutations identified in ~ one-third of squamous cell carcinomas Target with PI3K inhibitor, e.g. BYL719 DDR2 tyrosine kinase mutations identified in 4% of squamous cell carcinomas Shown to confer sensitivity to dasatinib

Design of a Disease-Specific Master Protocol Lisa LaVange, PhD FDA

Potential Leveraging of Control Subjects Leveraging control patients across multiple trials is possible if a neutral 3 rd party is running the trial A CRO/Coordinating Center must establish appropriate firewall procedures Active drugs would not be compared to each other Approval would be based on meeting pre-specified efficacy and safety criteria compared to SoC Which control patients are unique or shared may not need to be disclosed for analysis purposes Benefits to sharing control patients Reduced recruitment time Reduced trial costs

Example of Leveraging of Control Subjects Marker A/Drug A1 recruited/randomized 1:1 to Drug A1:SoC Marker A/Drug A2 s protocol approved to begin recruitment Randomization of Marker A patients changes to 1:1:1 Drug A1: Drug A2: SoC Use of common protocol with standard procedures, visit schedules, and CRFs allows control patients to contribute data to both trials A1 trial completes enrollment while A2 trial is still ongoing Randomization allocation reverts to 1:1 for Drug A2: SoC Shared controls that have completed follow-up in Drug A1 trial Data is unmasked for analysis of the Drug A1 protocol Data remains masked to Drug A2 trial personnel If necessary, data collection on A2 patients continues under the Drug A2 protocol

Scenario: Two markers with no marker overlap; two drugs target marker A. No marker overlap Advanced NSCLC patients Core needle biopsy Stratify by biomarker expression Divide in 2 portions: 1) Clinical trial biomarker analysis 2) Save for future analysis Drugs A1 B1 A+ Randomize 1:1 Drug A2 SoC A-B- A+B- A-B+ PFS; OS Randomize 1:1 Randomize 1:1 A+ Drug A1 SoC Drug B1 SoC SoC Randomize 1:1:1 Drug A1 trial arm finishes accruing PFS; OS PFS; OS Drug A1 SoC Drug A2 + Drug A2 PFS; OS PFS; OS

Potential Leveraging of Control Subjects Leveraging control patients across multiple trials is possible if a neutral 3 rd party is running the trial A CRO/Coordinating Center must establish appropriate firewall procedures Active drugs would not be compared to each other Approval would be based on meeting pre-specified efficacy and safety criteria compared to SoC It may not need to be disclosed for analysis purposes, which controls are unique or shared Benefits to sharing control patients Reduced recruitment time Reduced trial costs

Design of a Disease-Specific Master Protocol Jeff Abrams, MD NCI

ALCHEMIST Adjuvant Lung Cancer Enrichment Marker Identification and Sequencing Trial

Drug Biomarkers in Lung Adenocarcinoma TKI-sensitizing EGFR mutations: 10% in Western population Up to 50% in Asian population Enriched in: females non-smokers younger patients Multiple tests in clinical use No FDA-approved clinical assay ALK Rearrangement 5-7% in Western population FDA approved companion diagnostic: Vysis Break Apart FISH probe ALK-EML4 fusion

National Trial for Molecular Characterization of Early Stage Non-squamous NSCLC Eligibility: Diagnosis of NSCLC (non-squamous) Clinical stage I, II, or III deemed resectable Pathologic stage I, II, or III that: has been successfully resected adequate tissue available +/- local test for EGFR mutation or ALK rearrangement Patient Consent to allow donation of de-identified cancer information for research performance of central testing for adjuvant study referral 5 year follow-up: treatment and outcome contact regarding follow-up biopsy if cancer recurs (optionally) re-contact if no recurrence at end of study

Tissue Flow Consent & Register: A151216 Screening & Follow-up Protocol Pre-op Cohort SOP-driven FF/FFPE After resection, buffy coat Post-op Cohort Assess FFPE buffy coat CLIA-approved LAB EGFR mutation test (sequencing) ALK rearrangement (FISH) TCGA Genomic sequencing Transcriptome Methylation E4512: Erlotinib A081105: Crizotinib Other Adjuvant Studies

Data Flow Pre-op Cohort Consent & Register Post-op Cohort Collect local test info A151216 Registry Screen Neg. CLIA-approved Lab Marker Analysis EGFR activating mutation A081105 ALK+ E4512 5 year follow-up cohort TCGA Data Sequencing Database E4512 Trial Database A081105 Trial Database

Design of a Disease-Specific Master Protocol Karen Arscott, DO Lung Cancer Alliance

Design of a Disease-Specific Master Protocol Shakuntala Malik, MD FDA/CDER/OHOP

Master Protocol Concept Joint NCI Thoracic Malignancies Steering Committee & FDA Workshop Hyatt Regency Bethesda, Bethesda MD February 2-3, 2012. Strategies for Integrating Biomarkers into Clinical Development of New Therapies for Lung Cancer Objective: To bring together leading academicians, clinicians, industry and government representatives to identify challenges and potential solutions in the clinical development of novel targeted therapies for lung cancer Outcomes: Suggestion to develop Master protocols (by Dr Pazdur) for different stages of Lung Cancer

FDA Perspective Targeted drug development presents unique opportunities for personalized medicine Master protocols will Provide consistency of development approach regardless of intended target Better utilize limited resources (including patient resources) Bring safe and effective drugs to patients faster Discussions of trial design and endpoints with FDA to occur once protocol and statistical analysis plan are well developed/near-final

FDA Perspective FDA Drug approval will, however, depend on Integrity of data collected Results of the trials Drug effect isolated (clear attribution to drug) Results not only statistically significant but also clinically meaningful Toxicity of the drug. Available therapies at the time of approval Risk: Benefit Ratio

FDA-approved Therapy with Specific Targets in NSCLC Erlotinib (EGFR tyrosine kinase inhibitor) Bevacizumab (VEGF-A inhibition) Crizotinib (ALK inhibitor) Demonstration of specific molecular abnormalities in patient s tumor not required in FDA-approved indication for erlotinib and bevacizumab but is required for crizotinib indication

FDA Perspective Companion diagnostic assay/assay performance sufficient to reliably & reproducibly identify marker-positive population Exploratory studies for marker-negative population will need to be conducted to: Differentiate between prognostic vs predictive markers Support device/test kit claims CDRH should be involved at initial stages of assay development

FDA Perspective New drugs/indications for lung cancer will continue to be approved by FDA based on a demonstrated effect on a surrogate endpoint that is reasonably likely to predict clinical benefit in a population where there is unmet clinical need. Such approvals are likely to be based on relatively small trials; confirmatory trials will be required to confirm and characterize the actual clinical benefit.